1 |
On Friday 14 October 2011 06:11:40 Sergey Mironov wrote: |
2 |
> 2011/10/14 Mike Frysinger: |
3 |
> > On Thursday 13 October 2011 19:30:14 Sergey Mironov wrote: |
4 |
> >> 2011/10/14 Mike Frysinger: |
5 |
> >> > On Thursday 13 October 2011 15:15:40 Sergey Mironov wrote: |
6 |
> >> >> Hello. I have my arm-unknown-linux-gnueabi system which uses glibc |
7 |
> >> >> and busybox. Recently I realised that iconv program doesn't exist in |
8 |
> >> >> the tree. I've compared host's and target's glibc file lists and |
9 |
> >> >> found that arm's version doesn't install anything to |
10 |
> >> >> /usr/arm-unknown-linux-gnueabi/usr/bin. What can be the cause of |
11 |
> >> >> this? Small C program shows that iconv_open function returns error |
12 |
> >> >> code just like if there is no iconv at all. How shoud I make arm's |
13 |
> >> >> iconv work? |
14 |
> >> > |
15 |
> >> > see glibc's files/eblits/src_compile.eblit: |
16 |
> >> > is_crosscompile && sed -i '1ibuild-programs = no' config.make |
17 |
> >> > |
18 |
> >> > i imagine disabling that line will get you `iconv` |
19 |
> >> |
20 |
> >> Thanks, I will try it! Am I understand correctly that iconv itself |
21 |
> >> (library, not program) is a part of glibc and it is not possible to |
22 |
> >> exclude it during the build? So despite the lack of program I should |
23 |
> >> still have fully functional library. |
24 |
> > |
25 |
> > your understanding is correct, however ........ |
26 |
> > |
27 |
> > iconv() is part of glibc but it relies on all of the gconv shared libs |
28 |
> > found in /usr/$CTARGET/usr/lib/gconv/ to do its actually work. which we |
29 |
> > also incidentally delete when building the cross-compiler glibc (see |
30 |
> > src_install.eblit and look for "gconv"). |
31 |
> > |
32 |
> > you aren't the first person to find this behavior undesirable, and when i |
33 |
> > implemented it, it was more of "let's save space on things i don't think |
34 |
> > anyone will use". but if people are using it, then installing these |
35 |
> > things probably makes sense. |
36 |
> |
37 |
> Well, maybe I really want something strange. I guess, people often |
38 |
> install another, non-cross-compiled version of glibc on top of initial |
39 |
> one. I've tried to do so, but found that cross-emerge complains about |
40 |
> conflicts - it simply doesn't treat initial glibc as a package |
41 |
> installed on target. I saw 2 ways - either edit package.provided and |
42 |
> don't install new glibc or disable conflicts detection and overwrite |
43 |
> some target's /lib* and /usr/lib/* files. I've chosen first way since |
44 |
> I am afraid of getting a mess of two glibc's compiled with different |
45 |
> tools. But how do you (or other people) act in this situation? |
46 |
|
47 |
indeed, this is the current wonky behavior. i guess the thing to do would be |
48 |
two fold: |
49 |
- remove the disabling of tools/supplemental files so the cross-compiler |
50 |
glibc has all the same files as sys-libs/glibc |
51 |
- have crossdev automatically add a package.provided entry to the |
52 |
/usr/$CTARGET/etc/ tree for sys-libs/glibc |
53 |
|
54 |
> By the way, I also had to handcopy libstdc++.so from |
55 |
> /usr/lib/gcc/$CTARGET to /usr/$CTARGET/lib to make C++ programs work. |
56 |
> It is another thing which makes me thinking of installig full glibc on |
57 |
> top of cross-one. |
58 |
|
59 |
yes, the /usr/$CTARGET tree has no gcc files installed at all. in native |
60 |
installs, we've got /etc/env.d/ which adds the internal gcc paths via LDPATH |
61 |
to ld.so.conf, and then ldso at runtime finds libstdc++.so and friends. but |
62 |
gcc is not installed at all in /usr/$CTARGET. in this case, you should be |
63 |
able to emerge gcc into /usr/$CTARGET since none of the cross-compiler gcc |
64 |
files should be in there ... |
65 |
-mike |